7.5 Quality Assurance, Accreditation, and Statistics

7.5 Quality Assurance, Accreditation, and Statistics

Quality assurance is how a vascular lab proves its results are accurate and reproducible. The RVT exam tests both the statistics that quantify accuracy and the operational program that maintains it.

Diagnostic test statistics

Every test result falls into one of four cells when compared to a gold standard (truth):

From this 2×2 table:

• Sensitivity = TP / (TP + FN). The ability to correctly identify disease. A highly sensitive test has few false negatives — good for ruling out (a negative is trustworthy).
• Specificity = TN / (TN + FP). The ability to correctly identify the absence of disease. A highly specific test has few false positives — good for ruling in (a positive is trustworthy).
• Positive predictive value (PPV) = TP / (TP + FP). Of those who test positive, the fraction who truly have disease. PPV falls as disease prevalence falls.
• Negative predictive value (NPV) = TN / (TN + FN). Of those who test negative, the fraction truly disease-free.
• Accuracy = (TP + TN) / total. The fraction of all results that are correct.

Key exam point: predictive values depend on disease prevalence in the tested population, while sensitivity and specificity are intrinsic to the test. A carotid screening program in a low-prevalence population will have a lower PPV even with a good test.

Correlation against gold standards

A vascular lab validates its velocity criteria by correlating duplex results with a reference standard — historically catheter angiography, and now also CTA (CT angiography), MRA (MR angiography), or surgical/pathologic findings. Each lab should periodically confirm that its thresholds (for example, the PSV cutoffs for ≥50% and ≥70% carotid stenosis) match outcomes in its own patient mix, because scanner settings and populations differ.

Equipment quality assurance

A cracked transducer face or dead crystal elements produce a vertical band of dropout; phantom testing catches drift in measured distances and velocities before it affects patient results.

Accreditation

The IAC (Intersocietal Accreditation Commission) Vascular Testing accreditation is the recognized standard. It requires written examination protocols, appropriately credentialed technologists (such as RVT) and interpreting physicians, documented diagnostic criteria, correlation/quality-improvement review, and a maintained QA program. Many payers tie reimbursement to accreditation, which is why labs maintain these records.

Worked example

A lab tests 100 patients for ≥70% ICA stenosis. Duplex flags 30 positive; angiography confirms 25 of them and finds 5 more it missed. So TP = 25, FP = 5, FN = 5, TN = 65. Sensitivity = 25/(25+5) = 83%. Specificity = 65/(65+5) = 93%. PPV = 25/(25+5) = 83%. The lab uses these to decide whether to adjust its velocity thresholds.

Common traps

• Confusing sensitivity with PPV. Sensitivity is fixed by the test; PPV shifts with prevalence.
• Assuming high accuracy means high PPV in a low-prevalence screen — it may not.
• Treating phantom QA as optional — it is a core accreditation and safety requirement.
• Forgetting that each lab must validate its own criteria rather than borrowing another lab's cutoffs blindly.

Likelihood ratios and ROC curves

Beyond the basic 2×2 measures, the exam may probe how a lab compares competing velocity thresholds. The positive likelihood ratio equals sensitivity divided by (1 − specificity) and tells you how much a positive result raises the odds of disease; values well above 10 are strongly confirmatory. A receiver operating characteristic (ROC) curve plots sensitivity against (1 − specificity) as the threshold is varied; the closer the curve hugs the upper-left corner and the larger the area under the curve (AUC), the better the test discriminates.

Moving a PSV cutoff higher raises specificity but lowers sensitivity, and the ROC curve makes that trade-off visible so the lab can pick the threshold that best fits its clinical goal.

Continuous quality improvement and credentialing

Accreditation is not a one-time event. An IAC-accredited vascular lab runs ongoing continuous quality improvement (CQI): it tracks correlation between its duplex reports and confirmed outcomes, reviews discrepancies, holds regular case-review meetings, and documents corrective actions. Technologists must hold and maintain a recognized credential such as the RVT and complete continuing medical education, while interpreting physicians meet defined training and volume standards.

The exam treats these as the operational backbone of quality: a lab that scans well but cannot show documented protocols, credentialed staff, periodic phantom checks, and outcome correlation does not meet the standard. When a stem describes a quality gap, the most defensible answer is the one that closes the documentation and review loop, not the one that simply re-scans the patient.